Şirketlerin Yatırımcı Sunumlarının Metin Madenciliği ve Görsel Analiz Yöntemiyle İncelenmesi
Text Mining and Visual Analysis Method for Examining Companies' Investor Presentations

Summary/Abstract: The aim of this study is to reveal how Türkiye companies' investor presentations have transformed in terms of discourse and form using a data-driven approach. It examines the investor presentations of Turkish companies listed on the BIST 30 between 2015 and 2025 in terms of their textual, emotional and visual dimensions. A total of 649 presentations, underwent a comprehensive preprocessing phase. Topic modelling was performed using BERTopic, sentiment analysis was conducted using a BERT-based model adapted to Turkish, and a formal analysis was carried out based on visual criteria such as brightness and text density. The findings show that themes such as debt, cash flow, and cost pressure were the most prominent topics throughout the period, while themes such as renewable energy, production infrastructure, and digitalisation showed a clear rise, particularly after 2020. Themes related to financial difficulties are clearly expressed in negative language, while sustainability and investment-focused messages are framed in a stable and positive tone. Visual analysis reveals that investor presentations have shifted towards darker backgrounds and denser text blocks over time. Methodologically, the study makes a unique contribution to the literature as one of the first to apply an integrated text–sentiment–visual analysis approach to Turkish corporate documents.

• Details
• Contents